Dendritic cells (DCs) are the most potent antigen presenting cells and are efficient at inducing responses even to antigens to which the immune system has become tolerant. Thus for tumour immunotherapy, in which an immune response is raised against a tumour, the use of DCs may be ideal if they were made to present tumour specific antigens. DCs might also be used to present antigens derived from infectious agents, such as bacteria, viruses or parasites, providing protective or therapeutic vaccines for such diseases. However effective transfer of antigens into DCs for any of these targets has proved the greatest problem with this approach.
To provide a realistic chance of generating a therapeutic immune response against a tumour antigen or other disease related antigen, several conditions have to be met. Firstly, it is necessary to identify molecules whose expression is tumour or disease specific (or at least selective), and which can therefore serve as the target for an immune response. This task has proved very difficult for the majority of common tumours, but is solved in for example the case of cervical cancer by the presence, in most cases, of the viral oncogenes E6 and E7, and for other tumours, good candidate antigens are beginning to be identified. For example the MUC-1 gene product is over expressed in a number of tumours, including 90% of ovarian cancers. Various other tumour associated antigens have also been identified, any of which might be used in an immunotherapy treatment of cancer. These include gp100, MART-1 tyrosinase, MAGE, CEA, PSA and many others. Further tumor associated antigens will no doubt continue to be discovered over time. Secondly, following the identification of the antigen/antigens, it is necessary to deliver the antigens in an immunogenic form to the immune system. To generate the cellular immune response critical for tumour rejection, this means the proteins must either be delivered inside the cytoplasm of a host cell (a difficult task for high molecular weight protein antigens) or synthesized by the host cells themselves after gene delivery or DNA immunisation. Viral vectors which have been considered for this purpose include vaccinia, adenoviruses, or retroviruses.
The cell-type which is now widely recognised as providing the optimal immune stimulus is the dendritic cell (DC; see for example Girolomoni and Ricciardi-Castagnoli, 1997). Indeed the DC appears to be the only cell-type capable of stimulating a primary immune response in vivo, and moreover has even been shown to be capable of breaking established tolerance in certain circumstances. A number of groups are exploring the use of DCs in autologous adoptive immunotherapy protocols to stimulate immune responses against tumours in the hope that they may show a therapeutic effect. Such protocols involve culture and/or enrichment of DCs from peripheral blood, in vitro loading of DCs with antigen and reintroduction of the DCs to the patient or direct in vivo loading of DCs with antigen. However this approach has been hampered by the absence of efficient means by which to load these cells with antigens. Recent work has however shown that presentation of antigens by peptide pulsed DCs has produced anti-tumour responses in vivo (Celluzzi et al., 1996; Zitvogel et al., 1996). As regard to viral vectors, retroviruses do not give high efficiency gene delivery to dendritic cells (Reeves et al., 1996; Aicher et al., 1997), and in our hands, unlike work reported by others (Arthur et al., 1997), adenoviruses only give low efficiency gene delivery.
We have previously tested and reported that herpes simplex viruses (HSV) can efficiently infect and deliver genes to dendritic cells (Coffin et al., 1998; WO 00/08191). HSV has a number of advantages over other vector systems for this purpose, in that it can efficiently infect a wide variety of cell-types (including some very hard to infect with other vector systems e.g. Dilloo et al., 1997; Coffin et al., 1998), is easy to manipulate, and can accept large DNA insertions allowing the expression of multiple genes (reviewed by Coffin and Latchman 1996). Delivery of multiple antigens to dendritic cells ex vivo followed by re-introduction into the body or direct administration of antigens to dendritic cells in vivo may be particularly promising approaches to the treatment of some cancers and infectious diseases.
WO 00/08191 teaches that wild type herpes simplex viruses prevent antigen processing occurring in infected dendritic cells and that herpes viruses that either lack both functional UL43 and vhs genes or contain mutations that minimise immediate early gene expression are capable of efficiently infecting dendritic cells without preventing antigen processing occurring in the infected cells.